Mismatch negativity (MMN) is an event related potential (ER)) elicited by auditory stimuli that differ in one or more acoustic features from a train of previously presented stimuli. MMN has been postulated to represent the activity of an auditory comparator mechanism wherein a memory trace is set up for the repetitive stimulus and the occurrence of a different sound triggers neural activity signalling that change. MMN reflects activity underlying an early, preattentive and automatic accoustic processing stage, which makes it an ERP well suited as a probe of the auditory capabilities of infants and young children. Since the appearance of MNN demonstrates that the auditory system has the ability to detect a change in acoustic stimulation, it holds the promise of providing 1) a powerful electrophysiologic tool to study the limits of auditory discrimination processing during development and 2) a probe of deviatn auditory processing. The aims of this project are to determine the optimal stimulating conditions under which MMN is present in newborns, infants and children to tones and speech sounds; to define the spatiotemporal characteristics of the response as a function of stimulus and age; to test the predictive utility of MMN in a cohort of infants at risk for speech/language and cognitive dysfunction; and to assess the clinical correlation of MMN with acoustic processing dysfunction by testing a group of children with phonemic processing disorders.